796
Macromolecules 1999, 32, 796-800
Trichlorosilane Chemisorption on Surface-Modified Poly(tetrafluoroethylene) Bin Zhao, William J. Brittain,* and Erwin A. Vogler† Department of Polymer Science, The University of Akron, Akron, Ohio 44325 Received July 21, 1998; Revised Manuscript Received November 12, 1998
ABSTRACT: Modification of poly(tetrafluoroethylene) (PTFE) by trichlorosilane chemisorption to hydroxyl-modified (PTFE-OH) and silicate-modified (PTFE/SiO2) PTFE is reported. PTFE/SiO2 was prepared by exposure of PTFE-OH to SiCl4 vapor. XPS and contact angle analyses were consistent with the chemical composition of PTFE-OH and PTFE/SiO2. Chemisorption of trichlorosilanes from solution produced layers on PTFE-OH and PTFE/SiO2. The FTIR absorption intensity for these layers was consistent with monolayer formation. Monolayers with Br, CF3, CH3, CD3, and CHdCH2 terminal groups were prepared. Relative to deuterated hexadecyltrichlorosilane (DHTS) monolayers on PTFE-OH, higher contact angles were observed for DHTS monolayers on PTFE/SiO2 under a greater variety of deposition conditions. Reflectance FTIR indicated that the monolayers formed on PTFE-OH and PTFE/SiO2 exhibited crystalline-like order consistent with a well-ordered monolayer. Thermal annealing of the DHTS monolayers for several days at 125 °C did not produce any significant changes in water contact angles.
Introduction
Scheme 1
Poly(tetrafluoroethylene), PTFE, and related fluoropolymers from one of the most important families of engineering polymers and are known for their mechanical properties, chemical inertness, low coefficients of friction, very low surface energy, and high electrical resistance. These properties recommend fluoropolymers for numerous and diverse applications ranging from lowfriction/stick surfaces to medical devices. A familiar strategy in materials science has been to modify surface properties of a polymer to enhance interfacial properties such as adhesion or biocompatibility while, at the same time, maintaining favorable bulk properties. Consequently, surface modification of PTFE has been the subject of numerous studies. For example, Hubbell and co-workers1 recently summarized some of the literature methods for PTFE modification. The method of most interest to us is the work of Costello and McCarthy.2 These workers exposed PTFE to Me2SO solutions of the potassium salt of benzoin dianion. Electron transfer from the dianion to PTFE leads to loss of fluoride from PTFE and subsequent formation of unsaturated functionality on the surface. Several reactions of the reduced PTFE (PTFE-C) were used to introduce hydroxyl, amino, and carboxylic acid functionalities on the surface. In general, one of the most precise methods to create a chemically well-defined surface is self-assembly. Normally, this method is applied to the formation of selfassembled monolayers on silicate and gold substrates.3 The clear advantage of this technique is that telechelic polymethylene chains can be chemically anchored to a surface. The surface chemistry and energy of the resulting monolayer is thus largely defined by the terminal group. The extension of this technology, which has been principally applied to a metallic or oxide surface, to polymers would be useful given the widespread commercial applications. Toward this end, Whitesides and co-workers4 prepared monolayers on disordered polymer † Present address: Becton Dickinson Research Center, Research Triangle Park, NC 27709.
substrates by the self-assembly of trichlorosilanes on surface-modified polyethylene and poly(dimethylsiloxane). For polyethylene, the preparation of self-assembled monolayers (SAMs) involved three basic steps: surface oxidation with an oxygen plasma; formation of a silicate layer anchored to the oxidized surface by reaction with SiCl4; formation of the SAM by chemisorption of a trichlorosilane. We have successfully adopted this method for the surface modification of PTFE with trichlorosilanes. Results and Discussion Our experimental strategy for surface modification of PTFE can be summarized by the following sequence: reduction of PTFE with benzoin dianion; oxidation with diborane; reaction with SiCl4 (this step may be optional); trichlorosilane chemisorption (Scheme 1). We have used FTIR, XPS, and contact angles to characterize the chemistry of these surface modifications. Formation of PTFE-OH. The first surface-synthetic step is formation of PTFE with hydroxyl chemical groups on the surface. This was done according to the method of Costello and McCarthy.2 We obtained results in reasonable agreement with those of Costello and McCarthy. We treated PTFE with a Me2SO solution of the potassium salt of benzoin dianion. Subsequent treatment of the gold-colored films (PTFE-C) with a
10.1021/ma9811526 CCC: $18.00 © 1999 American Chemical Society Published on Web 01/14/1999
Macromolecules, Vol. 32, No. 3, 1999
Figure 1. Reflectance FTIR of PTFE (a), PTFE-C (b), PTFEOH (c), and PTFE/SiO2 (d).
THF solution of BH3 (20 h) and oxidation with a basic solution of peroxide produced film samples (PTFE-OH). XPS analysis was similar to the published results; for example, McCarthy reported an atomic composition ratio for C/O of 3.4/1 while we observed a C/O ratio of 2.9/1. ATR-IR analysis of PTFE-OH supports hydroxyl introduction (Figure 1); absorbances at 3500 and 1020 cm-1 are due to ν(O-H) and ν(C-O). Water contact angles for PTFE-OH were θa ) 64 ( 4° and θr ) 44 ( 3°. Costello and McCarthy reported a similar advancing angle but a much lower receding angle (θr ) 0°). We cannot account for the difference in receding angles for our samples compared to the literature; however, it is worthwhile to note in this connection that contact angles depend on several important experimental factors including roughness of the sample, methods of analysis, purity of water, and quality of starting materials which can vary from experimentalist to experimentalist. Formation of PTFE/SiO2. We reacted PTFE-OH with SiCl4 vapor to create a surface layer of silicate. The experimental method of Whitesides and co-workers4 was used. The PTFE-OH samples were held at a distance of 1 cm over liquid SiCl4 for 30 s; no effort was made to control ambient humidity. This procedure was repeated until the sample was fully wettable. XPS analysis of PTFE-SiO2 revealed atom % surface composition (excluding hydrogen) of 23.6% Si, 23.0% C, and 47.4% O, with the remaining 6 atom % comprised of trace quantities (10 nm) overlayer of silicate/monolayer on PTFE. This is consistent with our previous estimate of the SiO2 layer thickness. The XPS of SAMs all show an increase in C relative to the PTFE/SiO2. Reflectance FTIR Analysis of Monolayers on PTFE/SiO2. For SAMs prepared on flat substrates, reflectance and grazing angle FTIR has been an indispensable technique for characterizing the structural order in the monolayers. For example, the dichroic ratio can provide information about the average tilt of chains normal to the substrate surface. Allara and Swalen7 performed early work on the grazing angle FTIR spectroscopy of monolayers. Porter and co-workers8 studied the grazing angle FTIR of alkanethiol monolayers on gold. They studied the peak positions and intensities of the C-H stretching modes for CH3(CH2)nSH in the crystalline and liquid states and adsorbed on gold. For the SAM prepared from CH3(CH2)21SH, νa(CH2) ) 2918 cm-1 and νs(CH2) ) 2850 cm-1. These peak positions were similar to those for the crystalline state, 2918 and 2851 cm-1, but different from the liquid state, 2924 and 2855 cm-1. In going from the crystalline state (in KBr) to the liquid state, there are shifts of +6 and +4 cm-1 for the νa(CH2) and νs(CH2) modes. The peak frequencies for the 21-carbon thiol adsorbed on gold were crystalline-like and suggested the alkyl chains were fully extended in the all-trans conformation. In addition, the peak positions varied as a function of chain length; for the SAM prepared from CH3(CH2)5SH, νa(CH2) ) 2921 cm-1 and νs(CH2) ) 2852 cm-1, indicating a less crystalline-like order in the SAM. These results were analogous to results reported by Synder and coworkers9 for crystalline and liquid polymethylene chains. The application of reflectance FTIR to monolayers on PTFE/SiO2 must take into account the inherently disordered surface of the underlying PTFE. However, if it is assumed that SAMs are conformal with the surface morphology of the underlying PTFE, we can probe the local order of the chains comprising the SAM. We have used the positions of the asymmetric (νa) and symmetric (νs) C-D stretching modes of the CD2 group in monolayers prepared from DHTS. We chose DHTS so that we could unambiguously attribute the absorptions of interest to the SAM and not to adventitious contamination of the surface by hydrocarbon contaminants from the local environment. When we studied HTS layers on modifed-PTFE, the intensity of CH2 absorptions was not reproducible. Table 3 contains the peak position data for SAMs prepared from DHTS. Figure 2 displays the CD2 region
Macromolecules, Vol. 32, No. 3, 1999
Trichlorosilane Chemisorption 799
Table 3. Peak Positions for CD3(CD2)15SiCl3 (DHTS) C-D Stretching Modes in Liquid and Adsorbed on Silicate Surfacesa C-D peak position, cm-1 sample
νa(CD2)
νs(CD2)
liquid DHTS (neat) PTFE/SiO2/DHTS PTFE-OH/DHTS ATR/SiO2/DHTSb
2198 2196 2196 2196
2096 2092 2092 2091
Table 4. Effect of Thermal Annealing on Water Contact Angles (deg) of Surface-Modified PTFEa time at 125 °C, h
a Experimental error in assigning peak positions ) (2 cm-1. ATR/SiO2/DHTS: SAM prepared on ATR silicon crystal using DHTS.
b
a
Figure 2. Transmission FTIR of liquid DHTS (a); reflectance FTIR of DHTS monolayers on PTFE/SiO2 (b), PTFE-OH (c), and silicon ATR crystal (d). Absorption intensities have been normalized to facilitate comparison.
of the FTIR spectrum for the samples in Table 3. The absorption intensities have normalized to facilitate comparison. The peak positions of the monolayers prepared on PTFE/SiO2 and PTFE-OH are similar to those for the monolayer prepared on a silicon ATR crystal. We assume that the SAM prepared on the ATR crystal represents a well-ordered monolayer system based on literature reports on similar systems.10 Therefore, we conclude the monolayers on the PTFE/SiO2 or PTFE-OH have similar structural order compared to the well-ordered monolayer prepared on the silicon ATR crystal. The peak positions are comparable to those for monolayers of perdeuterated stearic acid, νa(CD2) ) 2195 cm-1 and νs(CD2) ) 2090 cm-1.11 For CD3(CD2)16COOH in CCl4, νa(CD2) ) 2200 cm-1 and νs(CD2) ) 2095 cm-1. One of the CD3 modes appears at 2216 cm-1 in the reflectance FTIR of DHTS monolayers on PTFE/SiO2 (Figure 2); it appears as a shoulder on the νa(CD2) peak. The ratio of intensities for the CD3 mode and the νa(CD2) peak (ACD2/ACD3) are 2.7 (PTFE/SiO2/DHTS), 3.0 (ATR/SiO2/DHTS) and 2.6 (PTFE-OH/DHTS) for the monolayers. For liquid DHTS, ACD2/ACD3 ) 1.3. As in foregoing discussion on peak positions, this ratio indicates that DHTS chemisorbed to modified PTFE leads
PTFE-OH θa
θr
PTFE-OH/DHTS θa
θr
PTFE/SiO2/DHTS θa
θr
0 17 43 68
Samples Prepared Using a 2 h Reduction Time 62 44 110 90 119 65 47 111 90 118 64 47 109 87 118 66 52 109 89 119
93 93 95 94
0 24
Samples Prepared Using a 0.5 h Reduction Time 67 49 113 91 119 77 61 111 89 120
96 95
Experimental error in contact angle measurements ) (3°.
to a monolayer with similar structural order compared to a SAM prepared on silicon. Thermal Reconstruction of Monolayers on Modified PTFE. Cross and McCarthy12 studied the thermal reconstruction of surface-functionalized poly(chlorotrifluoroethylene) (PCTFE). When samples were heated for days at 80-110 °C, surface-functionalized samples of PCTFE underwent surface reconstruction as evidenced by changes in contact angles and XPS data. For example, the water contact angle of hydroxyl-modified PCTFE increased +20° after annealing a sample for several days at 100 °C. The reconstruction was depth dependent and surface-chemistry dependent. We have investigated the possibility of the same phenomenon for DHTS monolayers on PTFE. Table 4 contains data for two sets of samples: one prepared by an initial reduction time (the time of treatment of PTFE by potassium benzoin dianion) of 0.5 h and the second set prepared by an initial reduction time of 2 h. Only the PTFE-OH samples showed significant changes in contact angles after annealing at 125 °C in air. The time of reduction presumably affects the depth of chemical modification but should not appreciably alter the subsequent modifications with SiO2/DHTS. The glasstransition temperature for PTFE is -73 °C.13 The changes observed for PTFE-OH are considerably smaller than Cross and McCarthy saw for PCTFE-OH although the trends are similar; in addition, the samples with shorter reduction times show the largest effects. Summary PTFE has been successfully modified by trichlorosilane chemisorption to hydroxyl-modified (PTFE-OH) and silicate-modified (PTFE/SiO2) PTFE. XPS results and contact angles are consistent with the anticipated surface chemistry. Evidence for monolayer formation is based on absorption intensities in reflectance FTIR analysis. We have used SAMs prepared on silicon substrates as an analytical standard to evaluate our results on surface-modified PTFE. The monolayers form equally well on both PTFE-OH and PTFE/SiO2, although higher contact angles were observed for DHTS monolayers on PTFE/SiO2 under a greater variety of conditions. Peak positions of the C-D stretching modes in DHTS monolayers are consistent with a crystallinelike monolayer suggesting structurally well-ordered monolayers. Thermal annealing of the DHTS monolayers did not produce any significant changes in water contact angles. The primary value of this surface modification technique is the opportunity to introduce specific functional groups onto the surface of PTFE with the same structural and chemical certainty that is normally associated
800 Zhao et al.
with the self-assembled monolayers on conventional silicate substrates. This is an avenue of research that we are currently exploring. Experimental Section Materials. PTFE was obtained from Bergoff/America as 1/32 in. thickness film, extracted with refluxing THF for 24 h and dried (
